92 BULLETIN OF WISCONSIN NATURAL HISTORY SOCIETY. VOL. 1, NO. 2. 
alive, together with the wasp larva, were transferred to a place in 
which they could be conveniently watched. None of the cater- 
pillars died until they were attacked. The larva ate all the food 
that was provided, the oldest one cocooning on the fourth, and the 
second one on the seventh of September. Of the third we have 
no record, excepting that the caterpillars had all been eaten on 
September eighth. 
We happened to be passing through our neighbor's grounds 
at nine o'clock on the morning of September fifth, and calling to 
ask whether there had been any more visits from the wasp, we 
learned that an Odyncriis had been seen making a mud partition in 
the horn on the day before, ^^^lile we were speaking she arrived 
and entered the mouth piece, where she remained for about ten 
minutes. When she departed we found that she had laid her egg, 
which we carried away with us, wishing to determine the length 
of the egg stage. This proved to be longer than that of any wasp 
that we had heretofore known, for not until the morning of Sep- 
tember ninth did the larva make its appearance, and then the 
hatching was accomplished in ^a manner new to us, the egg-skin 
1:)ursting and leaving its tenant free to crawl away. As a usual 
thing the egg changes into a larva imperceptibly, there being no 
sloughing off of the skin. 
Odynenis copra, then, first finds a suitable crevice and builds 
an earthen partition across the inner end, the earth being scratched 
up from some dry, bare spot and moistened in her mouth. In 
this her habits seem to be identical with O. nidulator, which is 
described by Fabre as building in natural crevices or in those 
excavated in the stems of plants by bees. She then lays her egg, 
very probably suspending it by a filament of web, as is done by 
O. nidulator and O. rcniforiiiis,! as well as by Emnenes.t Not 
until this is done does she gather the ten or tw^elve small cater- 
pillars which are to serve as food for her young. One cell being 
provisioned, others follow until the crevice is filled. The cater- 
pillars are not killed, nor even reduced to a state of decent im- 
mobility, since we were obliged to press wads of cotton into 
the tubes in which they were kept to prevent their wTiggling out 
of reach of the wasp-larvae. In the case of O. nidulator the cater- 
pillars, according to Fabre, are so severely stung that they are 
reduced to absolute immobility. Not the slightest response is 
given to stimulation, and had the observer only this test to depend 
upon they would be pronounced dead. In O. reniforinis, on the 
other hand, and to a somewhat lesser extent in Eumcnes, the con- 
(*) Fabre. Souveners Entomologiques, Ouatrieme Serie, 1891, p. 179. 
(i) Id. Nouveaux Souveners Entomologiques, 1882, p. 89. 
(I) Id., ibid, p. 74. 
